考点分析专家

Security checks across malware telemetry and agentic risk

Overview

This is an instruction-only education skill that analyzes user-provided syllabus or exam text and does not request system access, tools, credentials, persistence, or installation scripts.

Reasonable to install for syllabus or exam-point analysis. Users should avoid sharing confidential, copyrighted, or restricted exam materials unless they are allowed to provide them to the agent.

SkillSpector

By NVIDIA

Vulnerability Patterns

Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (1)

Vague Triggers

Medium

Confidence: 93% confidence
Finding: The activation condition is overly permissive because the skill can trigger on generic teaching-outline text alone, without an explicit user intent signal. In multi-skill or agentic environments, this increases the chance of accidental invocation, prompt hijacking by pasted content, or the skill intercepting documents the user did not mean to route here.

VirusTotal

63/63 vendors flagged this skill as clean.

View on VirusTotal